Graphite8five

Superbxxx
ok man. Hope it helps. You dont need a whole day, just one hour or 2.

 
I'd had a look at some 1080 Ti FTW teardown's and I think it's just a matter of separating the heatsink from the board to get access to the under side of the heatsink, but would I need to replace the thermal pads before reassembling or just whack it back together? 
 
Edit: Were your loose fins around this corner section of the heatsink?
 

You can reutilise the thermal pads just be gentle. You need to put some new tim on the gpu. My fins were loose there and on the left side os the picture.

Superbxxx
 
You can reutilise the thermal pads just be gentle. You need to put some new tim on the gpu. My fins were loose there and on the left side os the picture.

Are you sure you can reuse the thermal pads? Could cooling take a hit doing so? I only brought enough thermal pad to apply a fix really but's if it's worth waiting to apply new thermal pad I can order some more. I know the heatsink can come off the card without removing the backplate but I'm curious to see if I find what this guy did! https://www.youtube.com/watch?v=XD3DF5cFAJc 8:28
 

Superbxxx
 
I am sure that you can reutilise the thermal pads just be gentle. Just saw the video....that was strange. I dont believe that is the norm.....the other videos that i have seen dont have that plastic. but give it a go and report back please.

 
Hi Superbxxx,
 
I did take my card apart over the weekend. The FTW3 is not the easiest card to dismantle, there are four screws that essentially hold the heatsink to the baseplate/PCB and all of the teardown/waterblock installation videos I watched failed to even mention these. I started a thread on that here: https://forums.evga.com/FTW3-disassembly-confusion-m2724352.aspx
 
Anyway once I did get the heatsink separated from the baseplate/PCB unfortunately I could not find any loose heatsink fins like you suggested, I had a really good look but nothing appeared overly loose... While I had the heatsink separated I blew the high pressure Datavac duster over the fins to see if I could replicate the air pressure of the fans and the noise they make, but no, absolutely no type of resonance noise from blowing air through the heatsink... So I am of the belief now that the noise is somehow electrical related possibly caused by the card or the fans themselves...
 
I got to applying new TIM (NT-H1) an reassembling. I also removed the backplate's to check the thermal pads didn't have plastic still on them! But no mine did not. That guy must have been unlucky! I'm surprised something like that passed through QC also! The NT-H1 didn't provide me with any better temps which was a bit disappointing but I think I am going to do this again and this time apply Thermal Grizzly Conductonaut...
 
So no joy for me with the fan noise...
 
Edit: Hmm not so sure about Thermal Grizzly Conductonaut, bit risky that one, will have to think about it...
 
 

Graphite8five

Superbxxx
 
I am sure that you can reutilise the thermal pads just be gentle. Just saw the video....that was strange. I dont believe that is the norm.....the other videos that i have seen dont have that plastic. but give it a go and report back please.

 
Hi Superbxxx,
 
I did take my card apart over the weekend. The FTW3 is not the easiest card to dismantle, there are four screws that essentially hold the heatsink to the baseplate/PCB and all of the teardown/waterblock installation videos I watched failed to even mention these. I started a thread on that here: https://forums.evga.com/FTW3-disassembly-confusion-m2724352.aspx
 
Anyway once I did get the heatsink separated from the baseplate/PCB unfortunately I could not find any loose heatsink fins like you suggested, I had a really good look but nothing appeared overly loose... While I had the heatsink separated I blew the high pressure Datavac duster over the fins to see if I could replicate the air pressure of the fans and the noise they make, but no, absolutely no type of resonance noise from blowing air through the heatsink... So I am of the belief now that the noise is somehow electrical related possibly caused by the card or the fans themselves...
 
I got to applying new TIM (NT-H1) an reassembling. I also removed the backplate's to check the thermal pads didn't have plastic still on them! But no mine did not. That guy must have been unlucky! I'm surprised something like that passed through QC also! The NT-H1 didn't provide me with any better temps which was a bit disappointing but I think I am going to do this again and this time apply Thermal Grizzly Conductonaut...
 
So no joy for me with the fan noise...
 
Edit: Hmm not so sure about Thermal Grizzly Conductonaut, bit risky that one, will have to think about it...
 
 

Hi there.
 
At least you tried to solve it. If it is a fan you can experiment whith the precision x slow down/turn up one of the fan individualy( try to do it on idle....for obvias reasons) and see if that fan makes noise.
 
Can you try that?
 
Ps: The NT-H1 didn't provide  any better temps at all? that is strange.....do not put mutch of it, i use mx-4 whem i change my tim and i see benefits (maybe i have cards that had a lot of tim/not so grreat tim). In respect to the Thermal Grizzly Conductonaut beware that it is electrical conductive..... i personaly dont like but if you are willing to risk it go for it. 

Graphite8five

_Gir_
Obviously doesn't sound normal
Some thoughts-
Is this an open bench/case?
Did you check each fan individually?
If you pinch down on the cooler against the PCB does the noise change?
 

No it's not normal.
 
No, everything is installed inside a Corsair Obsidian 450D... Check each fan individually for what? They do hum individually from each other so it is not a rpm imbalance between the fans causing the noise... Do you mean apply pressure on the shroud towards the PCB? Slight pressure on the shroud doesn't seem to alleviate the hum. I will try to upload a video.
 
Edit: I captured the humming in this video and it is just the MEM fan running on it's own. I am adjusting the rpm in 1% increments where the tone changes slightly and sometimes adjusting to a rpm % where the hum drops away.
 
https://www.youtube.com/w...4&feature=youtu.be

 
MY CARD HAS THE EXACT SAME ISSUE!
Humming sound at certain rpm that is driving me nuts. I let EVGA Support know about it and send them 2 recordings upon request. But they fail to acknowledge it! I admit I don't have good recording equipment that picks those frequencies up very well. I guess it's hard to pick out the sound from a suboptimal recording. But for customers like us who sit next to it it's very easy to pick it out. Actually live in person the frequencies can easily be heard 2m away from my pc and it sticks in my head. Also it happens @ low rpm for me (around 1200rpm on the power-fan) when the cooler should be virtually silent. It's clearly a bearing defect on one fan. I've send a better recording to the EVGA support but a reply is still pending.
 
Even a German review site noticed the same issue on their review sample. (Source: https://www.computerbase.de/2017-10/evga-geforce-gtx-1080-ti-ftw3-elite-test/2/#abschnitt_fazit )
 
Translation:
"However, the left-mounted fan produces a quite unpleasant noise when changing the speed, which is not visible in the measurement series, but is immediately audible. The two other fans do not have this sound. If the speed remains constant, the noise characteristic is normalized. Thus, it is conceivable that it is a "Montagsmodell". According to EVGA, the phenomenon should not occur. In the next few days, a second copy will arrive at the editorial office in order to get to the bottom of the problem."
 
The answer I got from EVGA support however was ridiculous: "Fan sound is normal @100% fanspeed" Seriously? I had to laugh but then I got pissed. Not that they think I can't tell the difference between normal fansound at high rpm and an issue with the fan at low rpm after I owned plenty (EVGA) GPUs. No I got pissed that they didn't read my mail properly. I specified exactly at which rpm and fanspeed percentages it occures (32%). In the first recording I did 1% increments until the humming started and beyond until it disappeard and back down(between 25%-35%). In the second recording I let the normal fancurve do it's thing applied light load with gpu-z and let the fans spin up and back down (the humming is actually worse when the rpm are coming back down).
 
On a sidenote my card also hangs on boot every 1 out of 3 attempts (at gpu initialisation). And windows won't even load. So there is something off, too.
 
After hearing all the praise about the customer support it's actually my first time dealing with the EVGA support, so far it's been a mixed bag... Needless to say the card needs to be exchanged. 

Well the hang on boot does happen every time now. It is probably not driver related, because windows doesn't boot at all. My drivers and software are always up to date. I suspect a bios thing if any.
 
Man I feel you, new card same thing is frustrating... maybe something with their new fan batch or they changed the manufacturer, there could be many things. But it seems like one fan which could be any, is making this unpleasant noise. Hopefully they fix this for later batches of the card. Probably many users won't even notice this and think it's normal if  they don't have a reference. I'll get a new one and I'll report back. Fingers crossed.

Nellee
I have been experiencing the same issue with EVGA GEFORCE GTX 1080 Ti. As it is my rig has developed some other problems . I was going to take this card out and put it in a new rig but now I think I'll just get something different. The noise really is a too much. Lving in a flat with thin walls I would not be surprised if the neighbours can hear it. Anyway, if there was an easy solution I would take it but I think I will cut my losses and avoid buying similar. Can anyone recommend a simialr card that is not noisy ?

Need a better idea of what card you have first.

https://ru-clip.com/video/rDqHctG6lZA/how-to-ekwb-1080-ti-ftw3-waterblock-install.html
This is a good video of EK waterblock install